Yield Model Research Articles

Comparative Evaluation of CERES-Maize, WOFOST-Maize, and Ensemble of Models for Predicting Maize Phenology, Growth, and Grain Yield

ABSTRACT Crop simulation models are useful for determining the effects of different environmental and management factors on crop growth, phenology, and grain yield. In this study, two process-oriented simulation models, CERES-Maize, and WOFOST-Maize, were evaluated for anthesis date, maturity date, biomass yield, grain yield, harvest index, and maximum leaf area index (LAI-max), and their comparative performance was assessed. The field experiments were conducted using a split–split plot design with four sowing dates {25th May (D1), 4th June (D2), 14th June (D3) and 24th June (D4)} in the main plot, two cultivars (PMH1 and PMH2) in the sub plot, and three nitrogen levels (N1, N2, and N3) in the sub-sub plot at the Research Farm, Department of Climate Change and Agricultural Meteorology, Punjab Agricultural University, Ludhiana (Punjab), to record the dataset for the calibration (2020) and evaluation (2021) of these models. The statistical indices used for the evaluation of these models were R2, root mean square error (RMSE), normalized root mean square error (NRMSE), mean bias error (MBE), index of agreement (IoA), and coefficient of residual mass (CRM). The model output was evaluated for the individual and ensemble models (ENSEM). The results of the calibration process revealed that the NRMSE of the CERES-Maize model was below 9% for both cultivars, except for harvest index, whereas the NRMSE of the WOFOST model was below 10%, except for the grain yield of PMH1 and LAI-max of both cultivars. On the other hand, validation results revealed that the NRMSE of both models was below 9% for most parameters, except for LAI-max of the CERES-Maize model for both cultivars and grain yield and LAI-max of the WOFOST-Maize model of PMH2. In addition, the evaluation indices indicated excellent agreement between the observed and simulated results of CERES and WOFOST. However, the performance of the CERES-Maize model was slightly better than that of the WOFOST model, with a lower NRMSE (CERES: ranging from 3.83 to 6.02% and WOFOST: ranging from 3.58 to 8.63%), except for LAI-max. The maximum leaf area index was the least agreed-upon parameter for both models. The RMSE and MBE values confirmed these results. The most accurate results were obtained for the ensemble of models, which reduced the NRMSE of the anthesis dates, biomass yield, and LAI-max. The prediction accuracy of the ensemble of models was also improved for LAI-max by reducing the NRMSE to 5.80%. On the basis of these results, it was concluded that the CERES model is a more reliable tool for accurately estimating phenology and grain yield, whereas the ensemble of models improved the estimation accuracy of most parameters compared with any single model.

Integrating Climate Data and Remote Sensing for Maize and Wheat Yield Modelling in Ethiopia’s Key Agricultural Region

Integrating Climate Data and Remote Sensing for Maize and Wheat Yield Modelling in Ethiopia’s Key Agricultural Region

Enhanced multifaceted model for plasmon-driven Schottky solar cells with integrated thermal effects

This paper explores the development of an opto-thermal-electrical model for plasmonic Schottky solar cells (PSSCs) using a comprehensive multiphysics approach. We simulated the optical properties, power conversion efficiencies, and energy yield of PSSCs with varying nanoparticle (NP) configurations and sizes. Our spectral analysis focused on the absorption characteristics of these solar cells, examining systems sized 3 × 3, 5 × 5, and 7 × 7, with NP radii ranging from 10 to 150 nm. The study addresses a significant gap in solar cell research by presenting a novel multi-physics energy yield model for PSSCs decorated with gold nanoparticles (Au-NPs) on silicon absorbers. This integrated framework uniquely couples optical, electrical, and thermal responses for the prediction of global energy yield maps. Total spectral heat absorption was evaluated over a range of 300 nm to 1200 nm. This spectral heating was further deconvoluted into nanoparticle heating and thermalization heating in a silicon absorber. The findings indicated that the 5 × 5 NP array with a 70 nm radius enhances electrical performance, with the short-circuit current density (Jsc) reaching 11.54 mA/cm2—A 47% improvement compared to traditional bare silicon Schottky cells of 2 μm thickness. However, this electrical enhancement was also accompanied by a significant increase in heat generation within the nanoparticles, with thermal gains up to 182.5% relative to the bare silicon cells. This substantial rise in thermal energy highlights the critical need for advanced thermal management strategies to mitigate overheating and ensure the overall efficiency of plasmonic-enhanced solar cells. Enhanced energy yield maps confirm the model’s predictions, showing improved outputs globally, especially in sunny regions with potential annual energy yield gains up to 80 kWh/m2.

Effects of Sediment Content, Flooding, and Drainage Process on Rice Growth and Leaf Physiology of Early Rice During Heading–Flowering Stage

In recent years, there has been a notable increase in the frequency and intensity of floods and heavy rains, which has resulted in the frequent inundation of rice-growing areas. Flooding during the heading–flowering stages of early rice can result in significant yield losses. To elucidate the response of rice to sediment content, flooding, and drainage processes and their underlying mechanisms, a pot experiment was conducted to investigate the effects of sediment contents (S1: 0, S2: 0.10 kg m−3, and S3: 0.25 kg m−3), flooding time (F1: 3 days and F2: 6 days), and drainage time (D1: 3 days and D2: 6 days) during the heading–flowering stage on the oxidation resistance and grain yield of early rice in the Poyang Lake Region. At the same time, an experimental control group (CK) was set up with no sediment, no flooding, or no drainage treatment. The results showed that the flag leaf area of S1F1D2 treatment was diminished by flooding. The relative chlorophyll content (SPAD) reached its lowest value prior to drainage. The treatment of S2F2D1 showed the greatest decrease in SPAD value of 41.57%, which was only 53.88% of that of the control treatment. The activity of superoxide dismutase (SOD), peroxidase (POD), and the content of malondialdehyde (MDA) were observed to increase during the flooding period in comparison to the control treatment. The maximum values for these parameters were recorded at 5.68, 3.09, and 1.9 times higher than those of the control treatment, respectively. However, a decrease was observed after drainage. Furthermore, the occurrence of flooding during the early rice heading–flowering stage resulted in a notable reduction in the grain number per spike and the fruiting rate, consequently leading to a considerable decline in grain yields, with a decrease ranging from 31.81% to 69.96%. The findings indicate that flooding during the heading–flowering stage resulted in a reduction in early rice grain yield yet enhanced the antioxidant capacity of the leaves. Regression analyses indicated that a prediction model for the actual yield after flooding stress at the heading–flowering stage of early rice could be constructed using SFW as the independent variable. The findings of this study provide a theoretical basis for the formulation of a scientific and reasonable drainage scheme with the objective of reducing yield loss following rice flooding in the southern rice-growing region of China.

Development and validation of soil test crop response model for beetroot (Beta vulgaris) grown in ultisols of India

Soil Test Crop Response (STCR), a combined plant nutrient management system that enables to develop fertilizer prescription equations for balanced crop nutrition, higher yield, profitability, and better nutrient efficiency. Field trial was carried out on Typic Haplohmult soil of Nilgiris, Tamil Nadu during 2023-2024 by Implementing an Inductive combined with a targeted yield model. Field trial includes a gradient experiment with a green viz., Chenopodium album; a test crop experiment with beetroot (Hybrid Improved Crystal) and a validation experiment with beetroot. First, the fertility gradient was ensured by the biomass yield and soil fertility. Then, test crop experiment with beet root were conducted in the same field to derive the basic parameters viz., Nutrient Requirement (NR), contribution of nutrients from fertilizers (Cf), contribution of nutrients from soil (Cs) and contribution of nutrients from the Farm Yard manure (Cfym). Using the basic parameters, fertilizer prescription equations were developed based on Integrated Plant Nutrition System and validated. We found that 0.38, 0.29, and 0.46 kg of N, P2O5, and K2O, respectively, were required for producing one quintal of beetroot tuber under the integrated approach. Readily customized of fertilizer nutrient doses was developed for varying soil test values and desired yield targets of beetroot, for both inorganic (NPK) alone and NPK + Farm Yard Manure (FYM). The model was validated in the same soil series with the achievement of 40 and 45 tonnes of beetroot ha-1 with 100.9% and 96.9% of yield achievement, respectively. The Soil analysis crop response - combined Plant Nutrition System model proved that beetroot yield can be increased by 34.74%, in relation to the generally recommended dose This inductive method could save 37, 26 and 34 kg of Nitrogen, Phosphorus and Potassium, respectively when Nitrogen, Phosphorus, Potassium fertilizers are combined with 12.5 t ha-1 FYM as per soil test and targeted yield of beetroot.

Simulation of sugar beet yield

Mathematical simulation of agricultural crop yields is based on the physical principle of causal interactions balance in a closed physical system. The conditions for model verification are determined, allowing to obtain objective and reliable findings. It is noted that empirical equations representing the dependence of crop yields on crop-forming factors in the form of polynomials of any degree, obtained using the multiple nonlinear regression method, are valid only for the conditions of a specific field experiment. Using them, it is impossible to carry out theoretical generalizations that would allow developing these particular solutions into a generalized mathematical model. It is shown that mathematical model of yield, presented in multiplicative form, can include an unlimited number of yield-forming factors. Atmospheric precipitation was used as a factor characterizing the moisture supply for plants in case of no irrigation. The validity of the developed solution is confirmed by 13 years of data on the yield of sugar beet (NZ-type hybrid) cultivated in Belarus at the State Agricultural Institution “Molodechno Variety Testing Station”. The mathematical model of yield is presented in dimensionless form; all blocks of factors of this model are similarity criteria. This allows to compare the results of mathematical simulation of yield of any agricultural crop on soils with any agrochemical properties. Such an analysis cannot be carried out with the results of calculating yields using the formulas of private mathematical models of agricultural crop yields in the form of algebraic polynomials.

Evaluation of Statistical Models of NDVI and Agronomic Variables in a Protected Agriculture System

Vegetable production in intensive protected agriculture systems has evolved due to its intensity and economic importance. Sensors are increasingly common for decision-making in crop management and control of environmental variables, obtaining optimal yields, such as estimating vegetation indices. Innovation and technological advances in unmanned vehicle platforms have improved spatial, spectral, and temporal resolution. However, in protected agriculture systems, the use is limited due to the assumption of having controlled environmental conditions for indeterminate vegetable production. Therefore, sequential monitoring of NDVI is proposed during the 2022 and 2023 agricultural cycles using the Green Seeker® sensor and agronomic variables. This has created a database to generate predictive models of development and yield as a function of nutrient status. The results obtained indicate high significance levels for the development and NDVI curves in all phenological stages; in contrast to the yield predictive models, this is due to the maximum values (close to one) recorded for NDVI inside the greenhouse in comparison to the yield prediction obtained from the 18th week of harvest. Evaluating the models between NDVI and agronomic variables is not an index that offers certainty in predicting yield in indeterminate crops in protected agriculture production systems. This is due to the constant optimal development in response to controlled environmental conditions, nutrient status, and water supply inside the greenhouse, without the sustainability of yield, which decreases in the final stages of production until production becomes economically unprofitable.

A Seasonal Fresh Tea Yield Estimation Method with Machine Learning Algorithms at Field Scale Integrating UAV RGB and Sentinel-2 Imagery

Traditional methods for estimating tea yield mainly rely on manual sampling surveys and empirical estimation, which are labor-intensive and time-consuming. Accurately estimating fresh tea production in different seasons has become a challenging task. It is possible to estimate the seasonal yield of tea at the field scale by using the spatial resolution of 10 m, 5-day revisit period and rich spectral information of Sentinel-2 imagery. This study integrated Sentinel-2 images and uncrewed aerial vehicle (UAV) RGB imagery to develop six regression models at the field scale, which were employed for the estimation of seasonal and annual fresh tea yields of the Yunlong Tea Cooperatives in Yixiang Town, Pu’er City, China. Firstly, we gathered fresh tea production data from 133 farmers in the cooperative over the past five years and obtained UAV RGB and Sentinel-2 imagery. Secondly, 23 spectral features were extracted from Sentinel-2 images. Based on the UAV images, the parcel of each farmer was positioned and three topographic features of slope, aspect, and elevation were extracted. Subsequently, these 26 features were screened using the random forest algorithm and Pearson correlation analysis. Thirdly, we applied six different regression algorithms to establish fresh tea yield models for each season and evaluated their estimation accuracy. The results showed that random forest regression models were the optimal choice for estimating spring and summer yields, with the spring model achieving an R2 value of 0.45, an RMSE of 40.38 kg/acre, and an rRMSE of 40.79%. Similarly, the summer model achieved an R2 value of 0.5, an RMSE of 78.46 kg/acre, and an rRMSE of 39.81%. For autumn and annual yield estimation, voting regression models demonstrated superior performance, with the autumn model achieving an R2 value of 0.42, an RMSE of 70.6 kg/acre, and an rRMSE of 39.77%, and the annual model attained an R2 value of 0.47, an RMSE of 168.7 kg/acre, and an rRMSE of 34.62%. This study provides a promising new method for estimating fresh tea yield in different seasons at the field scale.

Modeling of pyrolysis oil yield from sawdust using fuzzy and Aspen Plus approach

Modeling of pyrolysis oil yield from sawdust using fuzzy and Aspen Plus approach

Construction and Optimization of Integrated Yield Prediction Model Based on Phenotypic Characteristics of Rice Grown in Small–Scale Plantations

An intelligent prediction model for rice yield in small-scale cultivation areas can provide precise forecasting results for farmers, rice planting enterprises, and researchers, holding significant importance for agricultural industries and crop science research within small regions. Although machine learning can handle complex nonlinear problems to enhance prediction accuracy, further improvements in models are still needed to accurately predict rice yields in small areas facing complex planting environments, thereby enhancing model performance. This study employs four rice phenotypic traits, namely, panicle angle, panicle length, total branch length, and grain number, along with seven machine learning methods—multiple linear regression, support vector machine, MLP, random forest, GBR, XGBoost, and LightGBM—to construct a yield prediction model group. Subsequently, the top three models with the best performance in individual model predictions are integrated using voting and stacking ensemble methods to obtain the optimal integrated model. Finally, the impact of different rice phenotypic traits on the performance of the stacked ensemble model is explored. Experimental results indicate that the random forest model performs best after individual machine learning modeling, with RMSE, R2, and MAPE values of 0.2777, 0.9062, and 17.04%, respectively. After model integration, Stacking–3m demonstrates the best performance, with RMSE, R2, and MAPE values of 0.2483, 0.9250, and 6.90%, respectively. Compared to the performance after random forest modeling, the RMSE decreased by 10.58%, R2 increased by 1.88%, and MAPE decreased by 0.76%, indicating improved model performance after stacking ensemble. The Stacking–3m model, which demonstrated the best comprehensive evaluation metrics, was selected for model validation, and the validation results were satisfactory, with MAE, R2, and MAPE values of 8.3384, 0.9285, and 0.2689, respectively. The above research findings demonstrate that this integrated model possesses high practical value and fills a gap in precise yield prediction for small-scale rice cultivation in the Yunnan Plateau region.

Monitoring Yield and Quality of Forages and Grassland in the View of Precision Agriculture Applications—A Review

The potential of precision agriculture (PA) in forage and grassland management should be more extensively exploited to meet the increasing global food demand on a sustainable basis. Monitoring biomass yield and quality traits directly impacts the fertilization and irrigation practises and frequency of utilization (cuts) in grasslands. Therefore, the main goal of the review is to examine the techniques for using PA applications to monitor productivity and quality in forage and grasslands. To achieve this, the authors discuss several monitoring technologies for biomass and plant stand characteristics (including quality) that make it possible to adopt digital farming in forages and grassland management. The review provides an overview about mass flow and impact sensors, moisture sensors, remote sensing-based approaches, near-infrared (NIR) spectroscopy, and mapping field heterogeneity and promotes decision support systems (DSSs) in this field. At a small scale, advanced sensors such as optical, thermal, and radar sensors mountable on drones; LiDAR (Light Detection and Ranging); and hyperspectral imaging techniques can be used for assessing plant and soil characteristics. At a larger scale, we discuss coupling of remote sensing with weather data (synergistic grassland yield modelling), Sentinel-2 data with radiative transfer modelling (RTM), Sentinel-1 backscatter, and Catboost–machine learning methods for digital mapping in terms of precision harvesting and site-specific farming decisions. It is known that the delineation of sward heterogeneity is more difficult in mixed grasslands due to spectral similarity among species. Thanks to Diversity-Interactions models, jointly assessing various species interactions under mixed grasslands is allowed. Further, understanding such complex sward heterogeneity might be feasible by integrating spectral un-mixing techniques such as the super-pixel segmentation technique, multi-level fusion procedure, and combined NIR spectroscopy with neural network models. This review offers a digital option for enhancing yield monitoring systems and implementing PA applications in forages and grassland management. The authors recommend a future research direction for the inclusion of costs and economic returns of digital technologies for precision grasslands and fodder production.

Climatic Condition–Based Comparative Study of Deep Learning Models for Yield Forecasting in Smart Agriculture

Climatic Condition–Based Comparative Study of Deep Learning Models for Yield Forecasting in Smart Agriculture

Creep model of bond-degradation in deep granite based on variable radius particle clump

The creep failure of rocks is related to its microstructure, external loading and time. A nonlinear yield model was introduced to describe the variation in the cohesion and friction angle with plastic strain and intergranular stress. The mechanical properties and creep characteristics of deep granite were obtained by indoor tests, and a variable radius particle clump model was constructed based on the particle flow method. The bond-weakening-friction-strengthening model was combined with the parallel bond stress corrosion method to establish the bond-degradation creep model of granite. The creep failure time, creep rate and tension and shear fractures number of the parallel bond stress corrosion model and the bond-degradation creep model were compared and analyzed to verify the applicability of the model. The fracture evolution law of deep roadway surrounding rock was studied based on the bond-degradation creep model. The results show that the rock failure characteristics and tension-compression ratio obtained by the variable radius particle clump modeling method are closer to the actual situation. Compared with the parallel bond stress corrosion model, the creep failure time of the bond-degradation creep model is shorter, more microfractures are generated during the failure process, and the numerical creep curves are more consistent with the test curves. The deep roadway vault shear failure and sidewall plate crack failure characteristics calculated based on the bond-degradation creep model are basically similar to the actual project situation. The bond-degradation creep model can better simulate the creep damage process of rocks under high stress, and is more suitable for analyzing the fracture evolution law of surrounding rock in deep hard rock cavern.

Mild drought conditions at the tillering stage promote dry matter accumulation and increase grain weight in drip-irrigated spring wheat (Triticum aestivum L.).

In order to elucidate the physiological mechanism of post-flowering assimilate transport regulating the formation of yields in arid regions and to provide technological support for further water-saving and high yields in the wheat region in Xinjiang, we conducted a study on the effects of different fertility periods and different degrees of drought and re-watering on the post-flowering dry matter accumulation and transport of spring wheat and the characteristics of grain filling. In two spring wheat growing seasons in 2023 and 2024, a split-zone design was used, with the drought-sensitive variety Xinchun 22 (XC22) and drought-tolerant variety Xinchun 6 (XC6) as the main zones and a fully irrigated control during the reproductive period [CK, 75%~80% field capacity (FC)], with mild drought at the tillering stage (T1, 60%~65% FC), moderate drought at the tillering stage (T2, 45%~50% FC), mild drought at the jointing stage (J1, 60%~65% FC), and mild drought at the jointing stage (J2, 45%~50% FC) as the sub-zones. The dry matter accumulation of the aboveground parts of wheat (stem sheaths, leaves, and spikes), the transfer rate and contribution rate of nutrient organs, the maximum filling rate (Vmax), and the mean filling rate (Vmean) increased significantly after re-watering in the T1 treatment, and decreased with the deepening of the degree of water stress. The 13C isotope tracer results also showed that the T1 treatment increased the distribution rate of 13C assimilates in the grain at maturity. Correlation and principal component analyses showed that grain weight was highly significantly and positively correlated with stem sheath, leaf, and spike dry matter accumulation, amount of nutrient organ post-flowering transports, transport rate, contribution rate, the onset and the termination time of the rapid growth period, Vmax, and Vmean, and stem sheath and spike dry matter accumulation had a direct effect on grain weight. While the two varieties performed differently among the treatments, both exhibited optimal performance in the T1 treatment. In conclusion, mild drought at the tillering stage (60%-65% FC) was the best model for water conservation and high yield of wheat under the conditions of this trial.

Energy yield framework to simulate thin film CIGS solar cells and analyze limitations of the technology.

This study presents a comprehensive evaluation of Copper Indium Gallium Selenide (CIGS) solar technology, benchmarked against crystalline silicon (c-Si) PERC PV technology. Utilizing a newly developed energy yield model, we analyzed the performance of CIGS in various environmental scenarios, emphasizing its behavior in low-light conditions and under different temperature regimes. The model demonstrated high accuracy with improved error metrics of normalized mean bias error (nMBE) ~1% and normalized root mean square error (nRMSE) of ~8%-20% in simulating rack mounted setup and integrated PV systems. Key findings reveal that the CIGS technology, while slightly underperforming in integrated, low-irradiance setups, shows comparable or superior performance to c-Si PERC technology in high-irradiance and high-temperature conditions. A significant focus of the study was on the low-light performance of CIGS, where it exhibited notable voltage losses. Our research highlights the importance of reducing the diode ideality factor for enhancing CIGS power conversion efficiency, particularly In low-light conditions. These insights provide a pathway for future research and technological improvements, emphasizing defect engineering, passivation strategies to advance the understanding and application of the CIGS technology.

A predictive model for biomass waste pyrolysis yield: Exploring the correlation of proximate analysis and product composition

A predictive model for biomass waste pyrolysis yield: Exploring the correlation of proximate analysis and product composition

МОДЕЛЮВАННЯ ВПЛИВУ КЛІМАТИЧНИХ ЗМІН НА ВРОЖАЙНІСТЬ ОЗИМОЇ ПШЕНИЦІ В ЛІСОСТЕПОВІЙ АГРОКЛІМАТИЧНІЙ ЗОНІ УКРАЇНИ (ЧЕРКАСЬКА ОБЛАСТЬ)

It is widely believed that climate change will reduce the yield of grain crops. In order to confirm or refute it, the authors carried out the mathematical modeling of the influence of climatic changes on the yield of winter wheat in some areas of the Cherkasy region, located in the forest-steppe zone of Ukraine. In the first stage, a mathematical model of the dependence of the yield of this crop on air temperature and rainfall was created. In the second stage, a mathematical model of winter wheat yield in one of the considered districts was generated, and calculations were made. The conducted modeling showed that the median of the winter wheat yield distribution function in the Drabiv district in the future period of 2030–2060 will exceed the historical values of this indicator for the period of 2005–2020 with a probability of 0.7. Moreover, with a probability of 0.4, the yield of winter wheat in the Drabiv district in the future period of 2030–2060 will exceed 50.7 c/ha. At the same time, the maximum yield in this area will not exceed 71.4 c/ha. Keywords: adaptation to climate change, crop productivity, quantile regression, statistical sampling, mathematical model.

Field-level sugarcane yield estimation utilizing Sentinel-2 time-series and machine learning

This work focused on developing a methodology for using machine learning (ML) approaches to establish a pre-harvest yield prediction model for sugarcane at field level by integrating time-series remote sensing imagery data with ML techniques. Ground truth agro data and thirty-one spectral vegetation indices were extracted from Sentinel-2 imagery and were considered for yield modeling. A two-level feature selection technique was used to determine the most significant variables that best correlated with sugarcane yield to predict yield in advance. Seven ML algorithms, including those based on regularization, decision trees, and ensemble methods like boosting, were used to predict yield. The approach achieved the highest R2 score of 0.73 and the lowest root mean squared error (RMSE) of 13.45 t/ha with random forest (RF) among the seven ML models tested. Furthermore, all feature selection procedures identified normalized difference red edge (NDRE), red edge chlorophyll index (RECI), and ratio vegetation index (RVI) as major yield-driving variables. The experiments during feature selection demonstrated the potential of red edge spectral bands in development of a reliable sugarcane-yield prediction approach. The RF model obtained using the proposed methodology outperforms the two baseline models developed using NDVI and GNDVI indices, with an improved RMSE of 16-18%.

Predictive Modelling of Local Currency Sovereign Debt yields using Machine Learning

This paper explores applying Statistical Machine Learning (ML) models to forecast local currency 10-year bond yields in 17 Emerging Markets (EM) economies, using Economic & Financial Indicators as feature parameters. The models considered include Random Forest (RF), Gradient Boosting (GB), Lasso Regression, Ridge Regression, and Linear Regression (LR). The performance is assessed using three key metrics: Mean Squared Error (MSE), R2 on the testing set, and R2 on Cross-Validation (CV). The results show that RF and GB models generally outperform traditional linear models, showing lower MSE and higher R2 values, indicating better accuracy in their predictions. By exploring the practical application of those models in the context of EM economies, this paper contributes valuable insights into the predictive modelling of sovereign debt yields, which is crucial for investors and policymakers. The findings suggest that ML models, with their ability to capture non-linear relationships, offer a more reliable approach to predicting 10-year bond yields in EM economies.

A New Yield Surface for Cemented Paste Backfill Based on the Modified Structured Cam-Clay

Cemented paste backfill (CPB) is a cemented void filling method gaining popularity over traditional hydraulic or rockfill methods. As mining depth increases, CPB-filled stopes are subjected to higher confining pressures. Due to the soil triaxial apparatus limitations, as the conventional method of triaxial testing on CPB, no confining pressures higher than 5 MPa can be applied to CPB over a range of curing time. This lack of data introduces uncertainty in predicting CPB behavior, potentially leading to an overestimation of the required strength. To address this, this study introduces a new testing method that allows for higher confinement beyond traditional limitations by modifying the Hoek triaxial cell to accommodate low-strength materials. This study then investigates the coupled influence of confining pressure and curing time (hydration) on CPB characteristics, specifically examining the impacts of different curing times and confining pressures on the mechanical and rheological properties of CPB. A total of 75 triaxial tests were conducted using 42 mm cylinder shape samples at five various curing times from 7 to 96 days, and applied at low and high confinement condition levels (0.5 to 30 MPa). The results reveal that hydration and confinement positively impact the CPB strength. The modified structured Cam-Clay model was selected to predict the behavior, and its yield surface was updated using the experimental results. The proposed yield model can be utilized to describe CPB material subjected to various curing and pressure conditions underground.